Chromosome folding and organisation across different organisms: a molecular dynamics study
Pereira, Maria Carolina
Polymer models have long been used to study the properties and behaviour of DNA, however the principles behind chromosome folding and organisation remain elusive. In this thesis we will analyse the contributions of different mechanisms driving genome compaction, such as macromolecular crowding, and interactions with different DNA-binding proteins. For this we will use Molecular Dynamics simulations of coarse-grained polymer models of both bacterial and eukaryotic DNA, together with methods of equilibrium and non-equilibrium Statistical Mechanics. Our study is motivated by recent experiments probing the compressional elasticity and dynamics of single bacterial chromosomes confined in a cylindrical pore, and by new high-throughput experimental techniques that capture the genome conformation in living cells. We start by looking at the properties of bacterial DNA. Our major contribution is the quantification of the effect of different compaction mechanisms on the DNA response to compression. We conclude that crowding proteins in particular strongly affect both the compression curves and the expansion dynamics. We also give evidence of a novel popping-off kinetic regime during expansion, where DNA-binding- proteins detach one by one leading to a slow unfolding dynamics. These results are robust with respect to changes in protein size and particle charge. We then turn to the study of the 3-D spatial organisation of human chromosomes. We conciliate two previously competing viewpoints regarding the mechanisms driving chromosome conformation in human cells. Thus, we show that transcription factors organise active and repressed chromatin, and establish long-range interactions leading to active/inactive domain phase separation, whereas chromatin loop extruding proteins (cohesin) are necessary to form domains within inert chromatin, which lacks binding sites for molecular bridges. We also show that a version of the model where chromatin loop extruders move diffusively rather than actively works equally well - this is important in view of single molecule experiments which have yet to find a motor activity in cohesin on chromatin fibres. Our model predicts the chromosome structure captured in experiments and the effect of various protein knock-outs. We finish by studying the nuclear organisation of fruit fly chromosomes and the subsequent formation of nuclear bodies. By modelling all chromosomes in the nucleus of an haploid cell, we are able to predict the structural and dynamical properties of the whole genome. We show that the formation of nuclear bodies is linked to genome reorganisation after mitosis (cell division). Our model predicts the dynamics and size distributions probed experimentally of such nuclear bodies. Importantly, we show that the large-scale chromosomal organisation is tightly dependent on the chromosomal conformation just after mitosis.